Available online at www.sciencedirect.

com

ScienceDirect
Procedia Engineering 99 (2015) 1275 1283

APISAT2014, 2014 Asia-Pacific International Symposium on Aerospace Technology,

APISAT2014

Multidisciplinary Design Optimization of Turbine Disks Based On

ANSYS Workbench Platforms
Qi Xiaodonga,*,Shen Xiulib
a

AVIC Commercial Aircraft Engine Co. Ltd., Shanghai, China

b
School of Jet Propulsion, BUAA, Beijing, China

Abstract
The design of aero engine turbine disks, which are working under high thermal and centrifugal loads, is an interactive and
multidisciplinary process that includes several disciplines, such as aerodynamics, structural analysis, mechanical design, and heat
transfer etc. Considering the actual aero-thermal-structure coupled environment, the main challenge for the designers is to
produce an optimum design that satisfies all the design criteria such as weight, life, efficiency and reliability.
Especially, problems of overweight and maximum localized stress have been always encountered for designers in the
preliminary design phase. Therefore, there exist contradictions for the design parameters from different disciplines, and then
multidisciplinary design optimization method could be a very valuable and efficient strategy to solve the problem.
In this paper, based on the platforms of ANSYS workbench software, the aero-thermal-structure coupled analysis of a high
pressure turbine disk had been conducted. Then, the parameters of width and height of the disk bore were selected as the design
variables, and multidisciplinary design optimization of turbine disk had been done. As a result, with the constraints on strength
criteria, the objective function of disk weight had been decreased by 6 percents which achieved the expected goal and
significantly reduced the design and research time.
2015
2014The
TheAuthors.
Authors.
Published
by Elsevier
Ltd.is an open access article under the CC BY-NC-ND license

Published
by Elsevier
Ltd. This
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility of Chinese Society of Aeronautics and Astronautics (CSAA).
Peer-review under responsibility of Chinese Society of Aeronautics and Astronautics (CSAA)

Keywords: Turbine disk, Multidisciplinary Design Optimization

* Corresponding author. Tel.: +86-02133367163;

E-mail address: 34040209@163.com

1877-7058 2015 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license

(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility of Chinese Society of Aeronautics and Astronautics (CSAA)

doi:10.1016/j.proeng.2014.12.659

1276

Qi Xiaodong and Shen Xiuli / Procedia Engineering 99 (2015) 1275 1283

1. Introduction
With the development of science and technology, the requirements of performance for aero engine are increasing.
As we all known, the most efficient way to meet the requirements is to increase the turbine inlet temperature. The
increase of turbine inlet temperature will inevitably raise the stress level of the turbine disc and thus affect the
reliability and life of the engine. So the new high-temperature resistant materials as well as the complex cooling
structure must be used to improve the turbine disc performance. The complex cooling structure can import a large
amount of air to cool the blade and the disc and in this process the temperature difference between the rim and bore
is reduced. Finally, it reduces the thermal stress. For cooling technology, on the one hand, the cooling effect must be
reliable, on the other hand, for the purpose of minimizing the harmful impact on the engine efficiency, it is
especially necessary that using the compressor air as little as possible on the premise of meeting the cooling
requirement. This is a complex aerodynamic and heat transfer process.
Whats more, turbine disc works at high speed environment withstanding the tremendous centrifugal force and
aerodynamic force caused by blades that involves complex aerodynamics, heat transfer, structural deformation, and
others. We need to find a method to do accurate and comprehensive analysis of the discs work status. That is
multidisciplinary design optimization.
The so-called multi-disciplinary design optimization (MDO) [1~2] is a new method proposed in recent years. The
main idea of MDO is to integrate the various disciplines knowledge and use effective design optimization strategies
and distributed computer network system to organize and manage the design process of complex systems. It is to get
the overall system optimal solution by making full use of the synergies caused by the interaction of disciplines.
Since from the MDO concept was put forward, a lot of fluid-thermal-structure coupled study about the turbine
disc and blade have been done [3~14], mainly for turbine blades about fluid-thermal or thermal-structure coupled.
Among them, [12~14] are relatively perfect in coupled problems. Reference [12] described an overview of the
loosely coupled model and a unified finite element method of solving fluid-thermal-structure problem. Reference [14]
proposed two kinds method of the three dimensional model when do multi-disciplinary analysis, brought about the
pass of data between the different disciplines meshed modal, and established a multidisciplinary method.
Based on the above research of the fluid-thermal coupled and thermal-structure coupled methods of blade, we
developed the fluid-thermal-structure coupled optimization method of the turbine disc.
2. Objects
In the concept design phase, the turbine disk design is usually based on two-dimension FEM analysis which could
be divided into two parts as is seen in Fig.1. Firstly, the design objective is the rim parts. Inputs are consisted of
blade pull, root chord, number of blades, flow path radius, etc. To minimum the weight of attachment, the rim width,
radius of live disk and basic shape parameters such as wedge angle, skew angle, the height of groove, etc. are
selected as the design variables. Secondly, the live disk is another part to be optimized. The objective function is the
weight of live disk, and the inputs are weights of blade and dead disk, rim radius, width of rim. Width of disk neck,
radius of disk bore, width of disk bore and the height of disk, etc. are recommended as design variables [15].
But in the preliminary design phase, considering the complex environments including heat transfer, mechanical
design and fluid analysis, three-dimension FEM analysis should be conducted to obtain the detailed geometry. By
multidisciplinary design optimization process, the engine performance could be further improved and the weight be
reduced.
In this paper, a high pressure turbine disk with 68 numbers of blade attachments was researched. The solid model
for strength analysis is seen in Fig.2 (a), and fluid model for fluid-thermal analysis is seen in Fig.2 (b).

Qi Xiaodong and Shen Xiuli / Procedia Engineering 99 (2015) 1275 1283

Disk concept design

process

Blade pull,
root chord,
number of blades,
flowpath radius

weights of
attachment,
rim radius,
width of rim

inputs

The attachment
design

Attachment
weight and
life?

outputs

inputs

No

The live disk

design

Disk weight and

life?

No

Gate 1:review on
the results

The end
Fig. 1: Turbine disk concept design process

Fig. 2: (a) Turbine disk solid model, (b) Turbine disk fluid model

1277

1278

Qi Xiaodong and Shen Xiuli / Procedia Engineering 99 (2015) 1275 1283

3. Multidisciplinary Design optimization

3.1. Design Process
Based on the ANSYS workbench software, the process of multidisciplinary design optimization in the
preliminary design phase is seen in Fig.3.
Parametric modeling of the turbine disk was performed by UG software. The feature parameters of disk and
attachments could be selected as the design variables for optimization. After the disk geometry was imported into the
Design Modeler (DM) module of ANSYS workbench, the feature parameters of disk could also be imported.
The fluid model was meshed by ICEM software in the Design simulation (DS) module. Then fluid-thermal
coupled analysis of turbine disk was conducted by ANSYS CFX with certain boundary conditions. As a result, we
got the temperature distribution of disks.
At the same time, in the Design simulation (DS) module, mesh, thermal and structure analysis of the solid model
were performed with the interpolated disk temperature from fluid-thermal coupled analysis above.
Finally, the optimization design was developed in the Design Xplorer (DX) module by updating the design
variables coming from UG software.
start

Design
parameters

UG modeling

Import
geometry

Material
parameters

Import
geometry
ICEM
meshing

mesh
Steady-state
thermal

temperature

CFX fluid-thermal
coupled

Static
structural
Design xplorer

Design variables;
State variables;
The objective function

Requirements?
No
Yes

end
Fig. 3: Preliminary multidisciplinary design optimization process

1279

Qi Xiaodong and Shen Xiuli / Procedia Engineering 99 (2015) 1275 1283

3.2. Fluid-thermal coupled analysis

The separate iteration method, also known as loosely coupled method is the most common method in fluidthermal coupled analysis of the turbine rotor. The whole calculation model will be divided into fluid domain and
solid domain through this method. Firstly, with a certain specified solid wall temperature conditions, the fluid
domain is analyzed to get the heat transfer coefficients of the solid domain. Secondly, the heat conduction equation
in the solid domain is solved to get the temperature distribution of the solid domain. Lastly, taking this temperature
distribution as boundary conditions, iterative process is carried out between fluid and solid domains until the result
trends towards convergent.
The fluid-thermal coupled model includes fluid domain and solid domain, as is shown in Fig.4. Fluid domain
includes disc cavity and internal cooling cavity of the root.
a

Outlet5

Outlet6

Outlet1

Outlet2

Inlet1

Outlet4

Outlet3

Inlet2

Fig. 4 :( a) Fluid domain, (b) Solid domain

The boundary conditions are inlet and outlet boundary, cyclic symmetry boundary and so on. The inlet and outlet
boundary conditions are shown in Table1~2. The solid domain boundary conditions are shown in Table 3.
Table 1. Inlet boundary conditions
Name

Inlet1

Inlet2

Pressure(MPa)

0.88P

Temperature (K)

1.08T1

T1

Table 2. Outlet boundary conditions

Name

Gas mass flow (kg/s)

Name

Gas mass flow (kg/s)

outlet1

Outlet4

1.73m

outlet2

Outlet5

4.42m

outlet3

1.92m

Outlet6

4.42m

1280

Qi Xiaodong and Shen Xiuli / Procedia Engineering 99 (2015) 1275 1283

Table 3 Solid model boundary conditions
Name

Blade root

Disk rim

Temperature (K)

T2

T2

*For simplicity, some of the data use a non-dimensional representation.

We set the conservation equations to the standard upwind method, and use four bands Rung-Kutta method and k turbulence model. Whats more, in this study we have assumed that the fluid is inviscid and the left and right fluid
walls are adiabatic, so in order to further simplify the calculation, the impact of speed is not taken into consideration.
Finally we got the heat transfer results of the turbine disk under steady-state flow. The speed streamlines and
temperature distribution results are shown in Fig.5.
a

Fig. 5: (a) Speed streamlines (Unit: m/s), (b) Temperature distribution (Unit: K)

3.3. Thermal-structure coupled analysis

The static structure analysis is calculated based on the fluid-thermal coupled results. By interpolating the fluidthermal coupled results as the boundary conditions of elastic-plastic stress analysis, we could get the stress
distribution of the turbine disk.
The meshed cyclic symmetry model of turbine disk is shown in Fig. 6(a).Boundary conditions contained axial
displacement constraints, cyclic symmetry boundary conditions, and the contact boundary between blade root and
disk.
The temperature data was transferred through GGI interface of CFX. The steady-state thermal analysis results are
plot in Fig.6 (b).

1281

Qi Xiaodong and Shen Xiuli / Procedia Engineering 99 (2015) 1275 1283

Fig.6 :(a) Meshed model of turbine disk, (b)Steady-state thermal analysis results(Unit: )

Through thermal-structure coupled calculation, the radial stress distribution of the turbine disk and blade is shown
in Fig.7. We can see that the maximum radial stress is at the web and the fillet of the third tooth region.
a

Fig. 7: (a) Static structure analysis results of disk (Unit: MPa), (b) Static structure analysis results of blade (Unit: MPa)

4. Results and discussion

The optimization design variables including width of disk bore and height of disk bore are listed in Table 4. And
the optimization results are seen in Table 5.
From the results above, we can see that the total weight of the disks is decreased by 6% through the proposed
fluid-thermal-structure coupled optimization design method. However the average hoop stress of disks is increased
by 7%, but still meets the burst margin requirements.
Table 4 Design variables
Variables
DS_W1
DS_H1

Name
width of disk bore
height of disk bore

Lower limit
1.000
0.429

Upper limit
1.334
0.715

1282

Qi Xiaodong and Shen Xiuli / Procedia Engineering 99 (2015) 1275 1283

Table 5 Optimization results
Variables
DS_W1
DS_H1
Weight(the objective function)
Maximum hoop stress
Average hoop stress
Maximum radius stress
Burst margin

Before optimization
1.190
0.619
1.000

0.799
0.758
1.35

After optimization
1.084
0.429
0.942
1.113
0.853
0.715
1.25

Difference
-8.91%
-30.7%
-5.87%
+0.11%
+6.70%
-5.73%
-7.40%

*For simplicity, some of the data use a non-dimensional representation.

And the response surface of weight and maximum hoop stress of disk to design variables has been also plot in
Fig.8.
a
b

Fig.8: (a) Response surface of disk weight (Unit:Kg), (b)Response surface of maximum hoop stress of disk (Unit:MPa)

5. Conclusions
Through all the research, this paper proposed multidisciplinary coupled analysis method of a high pressure
turbine disk model in the preliminary design phase. Based on the accomplishment of data transfer between different
disciplines and different meshed model, we finished multidisciplinary design optimization of turbine disk. The
optimization made the weight reduced by 6 percents which is of great significance. However, the locations of inlet
and outlet affect the airflow and cooling effect of the disc and we will do more research and optimization on them.
REFERENCES
[1] Wang Jingchao, Li Lizhou. Multidisciplinary design optimization development for turbine blade [J], Journal of aerospace power. 2005
[2]Jasmin Turcotte. Integration and multidisciplinary design optimization of a simplified gas turbine model using Perl and isight [J],AIAA 2004
[3] Hongjun LiNumerical prediction of fluid flow and heat transfer in turbine blades with internal cooling[R].AIAA94-2933
[4] Hylton,L.D., Analytical and Experimental Evaluation of the Heat Transfer Distribution Over the Surfaces of Turbine Vanes[R],NASA
Technical Report CR168015,1983
[5]Bohn,D.E,3D Internal Flow and Conjugate Calculations of Convective Cooled Turbine Blades with Serpentine-Shaped and Ribbed
Channels[C],ASME99-GT220, ASME Turbo Expo,Indianapolis,June,1999
[6] D.E. Bohn, Becker. 3-D Conjugate Flow and Heat Transfer Calculations of Film-Cooled Turbine Guide Vane at Different Operation
Conditions[C]. ASME Paper 97-GT 23.1997
[7]Han,Z.X., Simultaneous Prediction of External Flow Field and Temperature in Internally Cooled 3-D Turbine Blade Material[C],ASME 2000GT253,ASME Turbo Expo,Munich,Germany,May,2000
[8]Ding Shuiting. Fluid and heat transfer fo high pressure rotating turbine cavity of aero engine. Propulsion Technology. 2003:24(2), 118-121

Qi Xiaodong and Shen Xiuli / Procedia Engineering 99 (2015) 1275 1283

[9]Zhang Yanchun, Finite element analysis of structure stress for a gas turbine disc [J]. Gas Turbine Technology, Vol.20, No.3, 46~49, 2007.9
[10]Li Zhaoyang. Three dimensional transient temperature analysis and stress computation for a gas turbine disk [J]. Power engineering, 2006,
211-215
[11] Qi Wenkai, Chen Wei. Analysis of turbine disc response in thermal-mechanical loads [J]. Machinery Design & Manufacture, No.2, 89~91,
2004.
[12]Shen Wencai, Cao Yueyun. Thermo-Elastic-Plastic Analysis of Coupling Field and Touch Problem of Turbine Disc and Blade. Thermal
turbine. Vol.38,No.1,25~28,2009.3
[13] Li Na, Hong Huji. Finite Element Method Based on CBS for Fluid-Thermal-Structure Interaction [D]. Nanjing University of Aeronautics
and Astronautics, 2005.
[14] Wang Jingchao, Li Lizhou. Turbine blade flow-hot-solid coupling method analysis [C]. 13th engine structure strength vibration conference,
156~158,2006.
[15]J.G.Osmer, Pratt & Whitney Aircraft Group, Engine Life and Usage Methodologies for - Conceptual Design [J].AIAA-81-1651

1283